Every single cancer survivor fears relapse. While being free of the primary tumor after treatment is encouraging, it is emotionally devastating to step out of the clinic after a yearly checkup with a diagnosis of metastatic disease. Approximately 90% of cancer lethality is due to metastasis - the spreading of primary tumor cells to distant organs. Despite a dramatic increase in early screening methods and comprehensive treatment regimens for primary breast tumors, the overall mortality of patients who relapsed with metastatic cancer in the US has not changed in the past two decades. Clinical failure in treating metastatic tumors is largely due to the evolutionary nature of tumor metastasis. In order to survive and establish secondary tumors in a different tissue microenvironment at distant organs, disseminated tumor cells have to adapt (metastatic evolution) to the host environment. Clinically, metastatic tumors exhibit drastically different characteristics from their primary counterpart. Therefore, current cancer treatments, designed based on the features of primary tumors, are rarely effective. Optimal design of breakthrough anti-metastasis therapies relies on our in-depth understanding of the common and functionally important traits during the early metastatic evolution. In this proposed study, we will use breast cancer brain metastasis as model system and take state-of-the-art approaches to trace and analyze a single cancer cell within its precise metastatic microenvironment from the initial moment of metastatic colonization. This collaborative effort from a multidisciplinary team, including a cancer biologist, a computational biologist, a mathematician and a bioinformatician will allow us to: 1) depict the dynamics of gene transcription during the different stages of brain metastasis at single cell level; 2) visualize th behavior of a single colonized metastatic tumor cell and its interaction with the microenvironment through real-time intravital imaging and deep tissue imaging; 3) mathematically integrate molecular (RNA-seq), behavioral (intravital) and structural (deep tissue imaging) information to identify the common and functionally important early metastatic evolutionary traits during the breast cancer brain metastasis in its native tissue context. Shifting the current clinical treatment model relies on new in-depth mechanistic insight obtained through basic and pre-clinical innovative research. Utilizing cutting-edge sequencing and imaging modality in our proposed study, we will have the capacity to systematically dissect traits of early metastatic behavior and discover potential novel therapeutic targets for paradigm-shifting novel adjuvant therapy tailored to specific target metastatic evolution. The conceptual validation from our bold and pioneering attempt proposed in this study will serves as a blueprint for a new category of adjuvant therapies for effective anti-metastasis treatment.